TISSUE TROPISMS OF AAV VECTORS DEFICIENT

IN RECEPTOR BINDING

A Thesis

Presented in Partial Fulfillment of the Requirements for Distinction for the

Degree Bachelor of Science in the Division of Medical Technology

in the Allied Medical Professions of The Ohio State University

By

Jennette Kathleen Crumrine

*****

The Ohio State University

2005

Advisors:

Jeffery S. Bartlett, Ph.D., Research Advisor Department of Pediatrics and Department of Molecular Virology, Immunology, and Medical Genetics The Ohio State University

Approved by

Kathy Waller, Ph.D., Honors Advisor Associate Professor, Division of Medical Technology ▬▬▬▬▬▬▬▬▬▬▬▬ School of Allied Medical Professions Advisor The Ohio State University Division of Medical Technology

ABSTRACT

Adeno-associated virus (AAV) is a single stranded DNA virus with an

icosahedral 20-25 nm capsid comprised of three structural proteins, VP1, VP2, and VP3.

AAV is of great interest as a gene therapy vector for the treatment of cancer and genetic

diseases due to its ability to transduce both mitotic and postmitotic tissues, be produced at

high titers, and mediate long-term gene expression without pathogenicity. However, one

concern is that the AAV2 receptor, heparan sulfate proteoglycan (HSPG), is expressed on

a wide variety of human tissues. This can cause unwanted gene transfer and expression

when only a specific cell type needs to be transduced. Changing arginine amino acids

(a.a.) at positions 585 and 588, to alanines (R585A,R588A) has been shown to effectively

eliminate AAV binding to HSPG, while the insertion of small peptide epitopes (<20 aa)

that bind alternative cellular receptors has been shown to effectively retarget AAV

vectors to these new receptors. The amino acid sequence of the major capsid protein VP3

is contained within the two larger and less abundant capsid proteins VP1 and VP2, thus

mutations within VP3 are also present in VP1 and VP2. PCR-based site directed

mutagenesis of pACG2, which expresses all of the AAV capsid proteins, was used to

change the arginines at positions 585 and 588 within the VP1, VP2, and VP3 proteins to

alanines. Similarly, these arginine residues were also changed to alanines in a second

plasmid that expresses only VP1 and VP3 proteins. Different peptides that target the

vasculature endothelial growth factor receptor 2(VEGFR2) were then incorporated into

the R585A, R588A AAV2 capsid constructs. The ability of wild-type AAV2 vectors,

R585A, R588A AAV2 vectors, VEGFR2-targeted AAV2 vectors, and an AAV2 vector

containing the R585A, R588A mutation on only the VP1 and VP3 proteins to bind to

polystyrene beads coated with heparin and to HSPG-expressing Hela C12 cells were

assessed. Our preliminary data suggest that AAV2 vectors comprised of capsids with

R585A, R588A mutations were unable to bind to HSPG. The wild-type AAV2 was the

only virion that successfully bound to the beads and Hela cells. An affinity assay was

performed showing that with a competitor in 100-fold excess, AAV2 did not bind as well

to the Hela C12 cells similar to the heparan mutants. These preliminary findings suggest

that AAV2 does naturally bind to heparan, and that the rAAV2 mutants were deficient in

receptor binding. This can lead to successful elimination of unwanted gene transfer and

expression and the retargeting of the rAAV2 vector to tissues that were once thought

untargetable.

ACKNOWLEDGMENTS

Dr. Kathy Waller- Thank you for giving me the motivation and self-assurance to do the best that I could. Not only have you broaden my horizons in class, but also with your aspects on life. A positive attitude can always get you far in life. Continue you teaching style and never stop helping students find the right path to follow. Michele Suhie – Thank you for starting me on my honor’s curriculum and assisting in choosing a topic. The guidelines and goals that you gave me really helped overcome that wall of problems/stress. You were always there as a counselor to listen to all of my problems and lead me in the right direction. Dr. Jeffrey Bartlett- Thank you for having faith in me and allowing me to be a member of your research project. I am very grateful to have had this opportunity to work with you. I will always cherish the experience you have given me. Matthew Stachler- Thank you for your patience and insightful ideas that guided me in the right direction on numerous occasions in the laboratory. You were always there to help me out of a lot of sticky situations when no one else was. I wish you the best of luck in your future endeavors, for you have the possibility to go far. Sferra’s laboratory- Thank you for your gift of the competitor recombinant virus. I could not have completed my project on time if it wasn’t for the small amount of virus needed. Dr. Wilson - Thank you for your generous gift of paying for all of the Honor students posters. I would also like to thank my family who made all of this possible through their love, encouragement and sacrifices. You have always taught me to work hard and dream big. I owe all of my great accomplishments in my life to you, thank you. Finally I want to give thanks to my husband, Bryce, because he was always there for me regardless of the type of day I was having. When things went wrong, you helped me believe in myself and gave me the reassurance that everything would be all right. One thing you have taught me is that everything happens for a reason. Once you open up you mind, body, and soul you can finally listen to your spiritual side.

TABLE OF CONTENTS

Page

Abstract……………………………………………………………………………………ii

Acknowledgments………………………………………………………………………..iv List of Figures…..… ………..……………………………………………………….….vii

Chapters:

1. Introduction……………………………………………………………………..1-4

1.1 Gene therapy and vectors……………………………………………………...1

1.2 Adeno-associated virus……………………..………………………………1-4

1.3 Background of problem……………………………………………………….4

2. Review of Literature……………………………………………………………5-7

3. Materials and Methods………………………………………………………...8-15

3.1 Construction of AAV capsid mutant plasmid……………………………….8-9

3.2 Ligation and Transformation…………………………………………….……9

3.3 Screening and purifying individual plasmids……………………………...9-10

3.4 In vitro studies…………………………………………………………....11-12

3.4.1 Cell culture…………………………………………………………11

3.4.2 Production of recombinant AAV particles with modified capsid….11

3.4.3 Transfection………………………………………………………..12

3.5 Purification……………………………………………………………...……13

3.6 Virus Titers…………………………………………………………..……....13

3.7 Infectious Titers……………………………………………………………..14

3.8 Heparin-Binding Assay………..………………………………………...14-16

3.8.1 Poros Beads…………………………………………………….14-15

3.8.2 Cell Binding Assay………………………...………………………15

3.8.3 Affinity Binding Assay………………………………………….…16

4. Research Results………………………………………………………...……17-28

5. Conclusion……………………………………………………………….…...29-30

6. Bibliography……………………………………………………………….…31-32

LIST OF FIGURES

Figure Page

1 AAV Capsid………...…………………………………………………………2

2 Outer and inner surface map of AAV…………...…………………………….2

3 Satellitism of AAV with Adeno virus infection………...…………………….3

4 Western Blot of AAV capsid…………………………...……………………..5

5 Visual description of how the three plasmids are placed into the 293 cells and

recombine to produce virus…………………………………………..………12

6 Map of restriction digest for pACG…….……………………………………17

7 Photograph of restriction digest separated on 1.1 % agarose gel……………19

8 Graph of Heparin binding to POROS beads…………………………………22

9 Graph of Heparan Cell Binding experiment 1...……………..…………..…..24

10 Graph of Heparan Cell Binding experiment 2……...……....….………….…25

11 Graph of Hela Cell Binding with competitor………………………………...27

CHAPTER 1

INTRODUCTION

1.1 Gene therapy

Gene therapy is an experimental treatment being investigated to correct defective

genes that are responsible for disease development. Researchers are studying several

approaches of gene therapy to treat cancers and genetic diseases. One of the biggest

challenges in gene therapy is finding a suitable vector to carry DNA into tissues.

Problems for certain vectors in gene therapy include: provocation of immune response,

not invading both quiescent and dividing cells, lack of indefinite expression and

producible high titers (1). Leading candidates for gene therapy vectors are the adenovirus,

retrovirus, and recombinant engineered adeno-associated virus (rAAV). The adeno-

associated virus (AAV) may be a solution to the problems with gene therapy (1, 2).

1.2 AAV

The development of gene transfer vectors from the human parvovirus, adeno-

associated virus (AAV), has provided an efficient and effective way of delivering genes

into mammalian cells AAV genome is encapsulated as a single stranded DNA molecule

of 4.7 KB. The AAV capsid is comprised of three structural proteins VP1, VP2, and VP3

in an approximate ratio of 1:1:18 (3,4,5,6). The non-enveloped virion is icosohedral in

shape and is one of the smallest (20-25 nM diameter)

that has been described. Figure 1 demonstrates the size

and shape of the AAV vector courtesy of

http://www.scripps.edu/mb/goodsell/.

Figure 1: AAV capsid

Even though humans are a natural host for the wild-type AAV virus and at least

80% of the human population is seropositive, no known diseases have emerged from its

infection (7). An electron cryo-microscopy and image reconstruction of adeno-associated

virus type 2 is shown below in figure 2. Picture A is a view of the outer surface of the

capsid and B is the inner surface. The surface maps are colored in such a way to

demonstrate the depth of the surface, higher radii are shown in red and lower radii appear

blue (8).

Figure 2: Outer and inner surface map of AAV(8).

Advantages of AAV as a gene delivery vector include its ability to transduce both

mitotic and postmitotic tissues, be produced at high titers, and mediate long-term gene

expression without pathogenicity (2). AAV is a non-pathogenic defective member of the

human parvovirus family that requires co-infection of a helper virus to propagate itself

into the host genome. Figure 3

demonstrates the human adenovirus with adeno-

associated virus (arrow), a satellite virus.

AAV requires co-infection with a helper virus to

cause a productive infection (9). In the

absence of a helper virus, AAV will integrate

itself into the host, establishing latency

throughout the full life cycle of a cell (10).

Figure 3: Satellitism of AAV (arrow) with Adeno virus infection (9).

AAV is stable in its latent form and can be rescued to enter into a lytic phase upon

infection with a helper virus. In its latent form, AAV based vector is capable of long-

term, high-level gene expression even in competent hosts without showing any cellular

immune response or toxicity (11). This is practical because only a single injection of a

therapeutic gene is required; therefore, no re-administering is necessary. Furthermore, re-

administering the virus will increase the chances of the host to produce antibodies against

the gene vector, rendering it ineffective.

Another aspect of AAV is its ability to transduce both mitotic and postmitotic

tissues such as the lungs, intestines, neurons, muscle, and hematopoietic cells. Even

though AAV vectors seem to be a nearly ideal means of genetic therapy, major

limitations for clinical application still exist. Limitations of the AAV vector include its

inability to target all tissue types, redirecting viral vectors without unwanted transfer of

genes and expression, and incorporation of peptides into the viral capsid. The size of

ligands that can be inserted into the AAV capsid is limited due to its small capsid size of

20-24nm (1,2).

1.3 Background of problem

In gene therapy, adeno-associated virus is used to deliver a transgene into the

body. One of the major concerns with AAV2 as a gene therapy vector is the fact that its

receptor, heparan sulfate proteoglycan (HSPG), is expressed on a wide variety of human

tissues (1, 2, 3). This can cause unwanted gene transfer and expression when only a

specific cell type needs to be transduce. Further clinical development of AAV based gene

therapy will require developing AAV vectors with the ability to target specific tissue

types. This can be accomplished by the elimination of binding to HSPG and the

incorporation of targeting peptides directly into the AAV capsid. By eliminating the virus

from utilizing heparan receptors on its capsid protein through mutations, specific sites

can be targeted without unwanted gene expression. Point mutations and an addition of

small peptides can control how much AAV accumulates in organs of non-interest and

successfully retarget the virus to specific sites (3).

CHAPTER 2

REVIEW OF LITERATURE

The AAV capsid is composed of three structural proteins VP1, VP2, and VP3 in

an approximate ratio of 1:1:18 (3, 4, 5, 6). Figure 4 is a western blot

of the AAV capsids, VP1, VP2 and VP3 showing the concentrations

of each capsid. The capsids were detected by using monoclonal

antibodies that recognizes all three capsid proteins. Certain regions

of the AAV vector capsid protein can tolerate the insertion Figure 4: Western Blot of AAV capsid (8).

of exogenous peptides. One such example has been the insertion of an Arg-Gly-Asp (RGD) peptide following amino acid 588 in the normal

AAV capsid protein sequence. This RGD peptide allows the modified AAV vector to

invade via an interaction with cell surface integrin receptors and without the use of

HSPG. By adding the RGD to the virus, one does not need to worry if HSPG is present

(2). RGD insertion into the wild type capsid only enhances the chances of the virus to

bind to cells. RGD can still bind to heparan, thus in order to redirect the virus, an

elimination of HSPG receptors on the capsid is necessary.

Efforts to improve the packaging capacity of the recombinant AAV vectors are

still being investigated. Currently, only about 4.5Kb of DNA can be efficiently packaged

into an AAV particle. Understanding the primary steps of viral entry will also be

important in defining parameters for efficient gene delivery.

The tropism of AAV can be changed by genetically introducing a ligand peptide

into the viral capsid, redirecting the binding of AAV to other cellular receptors. By

introducing a mutant insertion into the capsid protein of AAV, we can retarget the

tropisms of the virus to cells that are normally resistant to AAV infection (12,13). AAV

has been mutated at numerous capsid protein subunits, one site being arginine 484 (R484)

and another being arginine 585 (R585). These mutations of the AAV capsid have been

shown to drastically reduce heparan binding. Tissue distribution in mice of the double

mutated AAV at R484 and R585 indicated marked infection reduction of the liver,

compared to infection with the wild-type AAV, but showed an increase of infection in the

heart (14). Various mutations sites have been undergone to the AAV capsid, but the

specificity and sensitivity of the mutated virus to bind heparan sulfate proteoglycan

in-vitro and in-vivo has never been demonstrated.

Current research has demonstrated that AAV can utilize different binding sites

other then HSPG to deliver transgenes into the body. Changing arginine amino acids

(a.a.) at positions 585 and 588, to alanines (R585A, R588A) has been shown to

effectively eliminate AAV binding to HSPG (3, 14, 15). The insertion of small peptide

epitopes (<20 a.a.) that bind alternative cellular receptors has been shown to effectively

retarget AAV vectors to these new receptors. The amino acid sequence of the major

capsid protein VP3 is contained within the two larger and less abundant capsid proteins

VP1 and VP2, thus mutations within VP3 are also present in VP1 and VP2. Addition of

small peptides (<20 a.a.) that bind specific receptors after amino acid 588 should

effectively retarget rAAV. Another approach involves the incorporation of larger proteins

(< 30 kDa) into the capsid into the minor capsid protein region VP2. The VP2 capsid

protein region is nonessential and can tolerate large peptide insertions at its N terminus.

This will minimize the disruption of the major structural protein, VP3 (6). What is

apparent is that there is a lack of correlation of heparan binding and infectivity of

some heparan-binding mutants. This means that HSPG might facilitate and enhance

AAV cellular interaction and infectivity, yet appears not to be necessary as a

receptor site (14). Further testing needs to be done to confirm that various point

mutations that reduce the binding and infectivity of the AAV capsid does not bind to

heparan sulfate proteoglycan. Once the tissue tropisms of AAV vectors deficient in

receptor binding have been investigated, the virus can be retargeted to tissues of interest

for gene therapy treatment.

CHAPTER 3

MATERIALS AND METHODS

The effects of various mutants that change receptor binding of AAV particles of

tissues in direct contact with blood and tissue fenestrations large enough for vectors to

enter (liver, lung, spleen, and bone marrow) was monitored. Different mutant viruses

were made with mutations in the AAV capsid that allowed infection independent of

HSPG and with peptide inserts for retargeting to tissues of interest. Dr. Bartlett’s lab has

developed methodologies for production and purification of AAV vectors discussed

below.

3.1 Construction of AAV capsid mutant plasmid

The plasmid pACG2 (16) was used as the template for all mutant constructions.

ACG contains the AAV genome, less the two viral inverted terminal repeats (ITRs), and

an ATG-to-ACG mutation of the Rep 78/68 start codon. This mutation has been shown

to increase recombinant AAV vector yield by attenuating Rep 78/ 68 synthesis (17).

Sites for mutagenesis were chosen on the basis of a computer-generated model of AAV

structure. Point mutations at the VP3 region on 585/588 where an arginine has been

change to an alanine were undertaken. This prevented the binding of AAV2 to heparan

and allows an insertion of new peptides to retarget the vector.

Another type of mutation is the insertion of peptides. Two PCR primers were

designed that contained the sequence of the desired insertion plus a unique

endonuclease restriction site flanked by 15 to 20 homologous base pairs on each side of

the insertion. The PCR products were digested with DpnI endonuclease to eliminate the

parental plasmid template. DpnI digests methylated DNA, which degrades the parent

plasmid. Mutant plasmids that were made in the PCR are unmethylated.

3.2 Ligation and Transformation

The plasmid were ligated together by T4 ligase and transformed into bacteria

(DH5-α), to allow screening of individual plasmids. The PCR purified ligated plasmid

was transformed by adding it to competent bacteria into an electroporation cuvette. The

cuvette is placed into an electroporation unit where two pulses of 1800 volts will be

passed through it. Time constants were monitored to assure recombination. The ligated

bacteria will be grown in liquid media (SOC) and then plated onto a selective media

plate with selection marker (AMP).

3.3 Screening and purifying individual plasmids

Colonies from plates were picked and grown in liquid media (LB or 2X-YT)

over night. The liquid bacteria will then go through a mini-prep procedure as

determined from QIAGEN. The bacteria will be purified and lysed to leave only DNA.

Once the plasmid DNA is purified it will be digested with specific restriction enzyme

endonucleases and ran out on a gel. This allows for screening of the plasmids to assess

if ligation of the peptides of interest were incorporated into the DNA plasmid. This step

can be repeated numerous times until a digestion that looks promising arise or changes

in mutations that should have been added demonstrate an extra band. An agarose gel is

made to the thickness desired of a one percent concentration. 1x TAE is used in

combination of powder agarose to make a gel. The liquid is heated and mixed until the

agarose dissolves and is cooled to about body temperature. Ethidium bromide is added

to the one percent solution in a ratio of 1µl per every 20 ml of mixture. Combs are

added to produce wells for the digested plasmid to enter. Loading dye is added to the

plasmid that will allow us to see the bands once ran out. A base ladder is added to the

sides to assess where the bands are migrating and whether or not the bands are the

correct base pair. Any mini preps whose digestion looks promising are sequenced to

confirm the proper mutant plasmid map. The sequences are read carefully to assure that

the mutant was added in the correct site with no other unwanted mutations in the capsid.

Sequences that were correct were grown in a larger amount of media (500ml)

and a maxi prep procedure from QIAGEN was preformed. The maxi prep procedure is

very similar to the mini prep, but only a larger amount of volume is used. The purified

plasmid will be placed in a spectrophotometer and read at wavelengths of 260 and 280

to find out the quantity and quality of the plasmid prep. 260 λ tells us the concentration

and the ratio of 260 to 280λ tells us the purity of the sample. All plasmids are stored in a

-20˚C freezer until further use. A portion of the liquid media with bacteria/plasmid

growth is mixed 50/50 with glycerol and stored in the -86˚C freezer until more plasmid

is needed. Once all the plasmids that are required for viral preps are prepared and

purified, tissue cultures can be started beginning the in vitro studies of the project.

3.4 In vitro studies

3.4.1 Cell culture

Low passage number Hela C12 cells, Hela cells, and HEK 293 cells (17)

were grown in Dulbecco’s modified Eagles medium (DMEM) supplemented with

10% fetal bovine serum, penicillin (100U/ml), and streptomycin (100U/ml) at 37 ˚C

and 5% CO2. Once the desired confluency was obtained, tests can be preformed or

the cells can be split using trypsin to aid in the breaking of the adhesion of the cells

to the bottom of the plate.

3.4.2 Production of recombinant AAV particles with modified capsids

To produce rAAV with mutant capsid proteins, 293 cells were transfected

with three plasmids. The three plasmids used were as follows: (1) pTR-UF5, which

contains the enhanced green fluorescent protein (eGFP) gene driven by the

cytomegalovirus (CMV) promoter and flanked by the AAV terminal repeats; (2) the

pXX6 plasmid, which supplies the adenovirus helper gene products in trans to allow

rAAV production in an adenovirus-free environment (16); and (3) the modified

pACG plasmid, which supplies the mutant capsid proteins. As a control, rAAV was

also prepared with unmodified pACG plasmid to make the wild type. The plasmids

were mixed at a 1:3:1 molar ratio. Plasmid DNAs used for transfection were

purified with a Quiagen (Valencia, CA) Maxi-prep kit according to the supplier

manual.

3.4.3 Transfection

Transfections were carried out as follows. 293 NC2 cells were split 48 hours

before transfection so that they could reach confluency around 75% at the time of

transfection. Four 10-cm plates were transfected at 37°C, using the calcium

phosphate transfection protocol standardized in Dr. Bartlett’s laboratory, and

incubated at 37°C (19). Forty-eight hours after transfection, cells were harvested by

centrifugation at 1200rpm for 10 minutes, resuspended in TMN (200mM Tris pH 8,

10mM MgCl2, 1.5M NaCl) and freezing and thawing three times released virus.

The crude lysate was clarified by centrifugation at 4200rpm for 20minutes and

treated with Benzonase (Sigma, St. Louis, MO) at 50-U/ml final concentrations at

37°C for 30 minutes. Virus was further purified by iodixanol step gradient (20).

Figure 5 demonstrates the three plasmids that are added to the 293 cells. After 48-72

hours, the virus is harvested out of the cells and then purified.

Figure 5. Visual description of how the three plasmids are placed into the 293 cells and recombine to produce virus.

3.5 Purification Once the virus has been harvested from the cells by a freeze thaw method, it

can be further purified through an iodixanol gradient. Different concentrations of

iodixanol is made and layered from highest concentration to lowest to a centrifuge tube

(Quick-Seal Ultra Clear). 60% iodixanol is placed onto the bottom of the tube, followed

slowly, trying not to mix, by decreasing concentrations (40%, 25%, and 15%). The

clarified lysate/virus is added to the top of the tube and then spun at 350,000 x g for 1

hour at 18˚C with a low brake. The tube is carefully removed and is aspirated with an

18-gauge needle at the 40% layer removing only the 40% solution. The virus will

migrate to the 40% layer due to weight/ density and proteins and salts will stay in the

25% layer.

3.6 Virus titers

The concentration of DNA-containing particles was determined by performing a

real-time (RT) PCR using a Perkin-Elmer (PE) – Applied Biosystems (Foster City,

Calif.) Prism 7700 sequence detector system. Briefly, 1µl of the purified rAAV stock

was diluted in 99µl of digestion buffer (10mM Tris-HCl [pH 8.3], 5 mM MgCl2, and 50

mM KCl ) containing Dnase I (350µg/ml). Samples were incubated at 37°C for 30

minutes, then 95°C for 10 minutes. Proteinase K was added to a final concentration of 1

µg/ml and the samples were incubated at 37°C for 30 minutes. Two and a half

microliters was used for RT-PCR in a Taqman reaction mixture and completed in

compliance as described by Clark (21). The only values that were accepted was within

the linear portion of a standard curve having a coefficient of linearity greater than 0.98.

The average RT-PCR titer was calculated from virus preparations assayed two times.

3.7 Infectious titers

Infectious titers of rAAV containing wild type or mutant capsids were

determined by gene transduction assay. Briefly, Hela C12 cells were seeded in 24-well

plates the day before infection so that they would reach about 75% confluency the next

day. Serial dilutions of wild type and mutant preparations were added to the cells in the

presence of Ad5 at a multiplicity of infection (MOI) of 3. The cells and viruses were

incubated at 37°C for 48 hours. Cells expressing eGFP transgenes were detected by

fluorescence-activated cell sorting (FACS) analysis. Infectious titers were also

determined in the absence of adenovirus with no differences in the relative titers of

mutants.

3.8 Heparin-binding assay

3.8.1 POROS Beads

The ability of mutants to bind heparin sulfate was assessed by precipitation with

heparin sulfate affinity resin (POROS). Briefly, 293 cells that were transfected and

contains wild type and mutant capsids were centrifuged at 4,600 rpm for 3 minutes

and the pellet was resuspended in a 500µl of lysis/binding buffer which contains

TMN as described before with the addition of 0.5% DOC, Benzonase (50 unit/ml),

and 10µl of POROS HE-20 resin. The samples were incubated for 15 minutes at

37°C to ensure complete cell lysis and then gently mixed at 25˚C for 2 hours. The

resin was washed three times with a wash buffer (12.5 mM Tris, pH 8.0; 0.5mM

MgCl2; 100 mM NaCl). The virus was eluted by adding 500µl of elution buffer (20

mM Tris, pH 8.0; 1mM MgCl2; 600 mM NaCl) and then a Taqman/DRP analysis

was performed as described previously. Controls included precipitation of vectors

with wild-type capsids and precipitation of vectors with unconjugated resin.

3.8.2 Cell Binding Assay

The ability of mutants to bind to the cell surface of Hela cells utilizing

heparan sulfate was assessed. Testing the ability of mutants to bind to Hela cells

versus heparin coated plastic polystyrene beads will give an efficient and more

realistic measurement of binding. Hela cells have been proven to express heparan

on their cellular surface in cell culture. The cellular binding assay is testing the

ability of the virus to bind to cells, most likely using HSPG. Hela cells were

harvested and centrifuged at 2,000 x g for 3 minutes and the pellet was washed once

with DMEM without serum and then twice with ice-cold binding buffer (DMEM

with 2mM glucose, 10 mM HEPES [pH 8] and 1% BSA). AAV vector/virus was

added at the highest MOI as possible. The samples were incubated for 60-120

minutes at 4 ˚C occasionally mixing gently. After incubation, the samples were

washed and resuspended in 1x TMN buffer with 0.5% DOC and Benzonase (50

unit/ml). The samples were incubated for 15 minutes at 37˚C to ensure complete cell

lysis and then spun to clarify the lysate. Taqman/DRP analyses on various serial

dilutions were performed.

3.8.3 Affinity binding assay

The ability of mutants to bind heparan sulfate on Hela cells was assessed in

combination with a competitor. Similarly to the cell-binding assay, viruses were

tested to determine if they could bind to the surface of the Hela cells. After cell

lysis, the lysate was analyzed by a Taqman/DRP titer to determine the amount of

virus in the solution. The viral mutants in our laboratory contain a CMV promoter

that drives the enhanced green fluorescent protein and allows the lab to determine

the concentration of virus in Taqman. A CMV probe is used in the Taqman/DRP

analysis procedure that will pick up any amount of virus in a solution tested. The

competitor virus was a gift from Sferra laboratory, which does not contain the CMV

promoter. Thus the Taqman CMV probe will not pick up the virus. The virus

CC119 from Sferra’s lab is a rAAV2 Eflα eGFP at a concentration of 7.6x 1012

DRP/ml. The affinity assay was performed following the cell-binding assay, only

when the viral mutants were added, as was the competitor. The competitor was

added in an excess of 100-fold. The procedure was continued exactly as the cell-

binding assay.

All methodologies used in this study are well established in Dr. Bartlett’s lab at

CRI. The appropriate AAV plasmids, antibodies, and other reagents were all on

hand or commercially available. Dr. Bartlett is an assistant professor in the

Department of Pediatrics and Department of Molecular Virology, Immunology, and

Medical Genetics at The Ohio State University.

CHAPTER 4

RESEARCH RESULTS

Point mutations and peptide insertions were successfully added to the pACG2

plasmid. PCR-based site directed mutagenesis of pACG2, which expresses all of the

AAV capsid proteins, was used to change the arginines at positions 585 and 588 within

the VP1, VP2, and VP3 proteins to alanines. The map in figure 6 shows some of the

restriction sites. Endonuclease enzymes cut the plasmid in desired places and inserts/

mutations were added.

Figure 6. Map of restriction digest for pACG.

The vector plasmid with a 585/588 point mutation of an arginine to an alanine of

the VP1, VP2, and VP3 capsid was screened, made, and peptide inserts were

successfully added. Peptides inserted that are of interest were a vascular endothelial

growth factor receptor 2 (VEGFR2) peptide. This would retarget the AAV virus to the

vasculature, which is of interest in our laboratory. Wild-type AAV does not transduce

the vasculature very well. Many potential benefits by targeting the vasculature can be

envisioned for gene therapy of cancers and cardiovascular diseases. Peptides include:

VEGFR2-A (VR2-A), VR2-B, VR2-C, VR2-D, and VR2-E. The peptide sequences for

the inserts were added, screened, and made. The inserted peptides will be displayed on

all three viral capsid proteins (VP1, VP2, and VP3). This allows for a large number of

peptides to be displayed, but only a limited size of peptides can be added. Below is the

actual peptides inserted into the 585/588 mutated pACG capsid.

VEGFR 2 A (VR2-A) 5’ CC GGT TGC GCC ACC TGG TTG CCC CCC CGC TGC GGC 3’ A ACG CGG TGG ACC AAC GGG GGG GCG ACG CCG AAT T

VR2-B 5’ CC GGT TGC TTG CCC CCC AAC CCC ACC AAG TGC GGC 3’ A ACG AAC GGG GGG TTG GGG TGG TTC ACG CCG AAT T

VR2-C 5’ CC GGT TGC TAC GCC ATC ATG CCC TTG GTG TGC GGC 3’ A ACG ATG CGG TAG TAC GGG AAC CAC ACG CCG AAT T

VR2-D 5’ CC GGT TGC ATG CAC TCC GAC ATG CAC GCC CCC GTG TCC GAC ATC TGC GGC 3’ A ACG TAC GTG AGC CTG TAC GTG CGG GGG CAC AGG CTG TAG ACG CCG AAT T

VR2-E 5’ CC GGT TGC CAC ACC ATG TAC TAC CAC CAC TAC CAG CAC CAC TTG TGC GGC 3’ A ACG GTG TGG TAC ATG ATG GTG GTG ATG GTC GTG GTG AAC ACG CCG AAT T

The pACG plasmid was cut with enzymes BST 98 I and AGE I for the above

VR2 peptides. The peptides VR2A-E were ordered with matching sticky ends shown

above as the underlined peptides, to match the cut plasmid. The peptides were annealed

into the capsid of the plasmid with the point mutations of the arginine amino acids (a.a.)

at positions 585 and 588, to alanines (R585A, R588A), and then ligated back together in

a PCR unit. The ligations were transformed into dsDNA, grown, and mini prep to look

for clones. The mini preps were digested with the enzymes BST 98I and VSP I, and ran

out on a gel. Figure 7 shows a few of the VR2 samples that were ran out and examined.

Figure 7: Photograph of restriction digest separated on 1.1% agarose gel. The vectors were cut using BST 98 I

and VSPI enzymes in the restriction digest.

The mini preps with arrows are the bands that looked promising based off of the ladder,

and were further sequenced to confirm the mutation.

Mutations of the AAV2 capsid were also undertaken in only the VP1, VP3

portion. VP2 was removed so larger peptides could be inserted into the plasmid, but

only displayed a limited amount of times. One of the pitfalls of the AAV capsid is its

small size, so it cannot withstand large peptide insertions. What is apparent is that the

VP2 capsid is not necessary for viral production. An AAV2 HS- VP1, 3 regions can be

combined with a separate VP2 region and still make virus. The VP2 protein can

withstand large insertions of peptides into its N-terminal region, like the VEGF

receptor. A larger VEGF insertion into the VP2 region has been under investigation, but

is still to be made. This would knock out heparin binding and allow the rAAV vector to

be retargeted to an area of interest. Once we get the VP2 regions with larger peptides to

incorporate properly back into the capsid protein, we can target areas in the body with

AAV2 that has never been targeted before.

HS-RGD plasmids were also made where an insertion of an Arg-Gly-Asp

(RGD) peptide following mutation of R585A, R588A in the normal AAV capsid

protein sequence. This RGD peptide allows the modified AAV vector to interact with

cell surface integrin receptors and without the use of HSPG. Current research has

demonstrated that the RGD peptide can successfully enhance the binding capacity of

rAAV regardless if HSPG was present. Unfortunately the RGD insertion in the

laboratory is not working and is currently under investigation. The HS- RGD AAV does

not bind heparan but it also does not infect cells either.

All of the mutated plasmids were transfected into 293 cells, harvested, purified

and tested for its ability to infect cells. Viruses were DRP titered to determine their

concentrations. A 24 well plate with 90% confluent Hela c12s cells were infected with a

MOI of 1000 of cell lysate virus of wild-type AAV, HS- AAV, and a control with no

virus were tested. Adenovirus was added to half of the wells with an MOI of 10 to show

the increase infectivity when a helper virus is present. The wild-type virus infected the

Hela c12 cells the highest with the helper virus and also without a helper virus. Only a

few cells were infected with the HS- AAV with helper virus and none without the

helper virus. The control was absent of infectivity with and without the helper virus.

The mutated viruses were tested for its ability to bind HSPG.

The ability of wild type AAV2 vectors, and heparan mutants AAV2 vectors to

bind to polystyrene beads coated with heparin was assessed. Every assay was performed

in triplicates to assure no variations between tubes. A wild type AAV2 vector, a heparan

mutant AAV2, and a control were run with every test. Variations in the original

protocol had to be undertaken due to unstable results. The original protocol called for

viral preps as crude lysates and after numerous repetitions of the assay, variations had to

be undertaken. Crude lysates is the viral particles straight out of the harvest cells. The

viral preparations are very “crude” in that cell debris, proteins, and salts are also present

in the lysate, which was a possibility for the unstable results. All of the crude lysate

viral preparations undergone iodixanol gradient purification which gave more reliable

results with lower standard deviations between samples.

Viruses that were tested for heparin binding included: wild-type AAV2, R585A

R588A AAV2 vectors, VEGFR2-targeted AAV2 vectors, AAV2 vector containing the

R585A, R588A mutation on only the VP1 and VP3 proteins, RGD AAV2 vectors, and

R585A R588A RGD AAV2 vectors. The RGD vectors did not behave as envisioned,

bringing on another twist onto why the vector was not infecting cells. RGD AAV2

vectors used to work in the lab by binding to cells whether heparan was present or not.

This would aid the AAV2 vector in binding when heparan sulfate proteoglycan may not

be present in the body. Unfortunately, the RGD vector recently stopped working and

investigation has been undertaken to troubleshoot the dilemma. The entire RGD vector

runs on the heparin resign were ignored due to the inconsistencies. All of the viral

vectors were allowed to bind to the polystyrene beads coated with heparin, washed, and

eluted off. While over 63 percent of the wild-type AAV2 was recovered in the elution,

less than 1 percent of the heparan mutants were found in the elution. Figure 8 is a graph

of the binding for an assay with wild type AAV2, R585A R588A AAV2 VR2B, and

AAV2 R585A R588A VP1, 3 plus VP2.

Heparin Binding

0

10

20

30

40

50

60

70

80

90

AAV2 R585A R588A AAV2VR2B

AAV2 R585A R588AVP1,3+VP2

Perc

en

t V

iru

s Elu

ted

Figure 8: Heparin binding using polystyrene beads coated with heparin and

iodixanol purified virus. Error bars indicate the standard deviation between the triplicate samples of each virus. Wild-type AAV2 bound 63%of the time while the R585A R588A AAV2 mutants bound less than 1%.

The above results demonstrated that I successfully knocked out heparan binding

in the AAV2 vectors. The viruses were next tested for their ability to bind to heparan

sulfate proteoglycan in a more realistic measurement. Cell binding of the viral preps to

Hela cells were examined for heparin binding. First experiments were run with 2.5 x 105

cells, which produced a weak pellet and were barely visible for washing. After the first

run, it was decided to start with 5.0 x 105 cells for a larger, visible pellet. All cells were

spun at 2,000 x g for three minutes and then washed with ice-cold DMEM without

serum one time. Next the cells were washed two times with a Binding Buffer and then

resuspended in a minimal volume. The AAV vectors were added to the appropriate

tubes at the determined MOI. The MOI for the first run with only 2.5 x 105 cells was

10,000 and MOI for second run was 5,000, resulting in 2.5 x 109 total virus added to all

test tubes in both runs. All tubes were run in triplicate alongside a control tube without

virus, consisting of just Hela cells. Due to unwanted background pickup caused by the

Taqman machine noticed in previous experiments with the polystyrene beads heparin

binding, serial dilutions were preformed on the viral samples to determine if any

significant binding occurred.

In the first experiment the viral samples were diluted 1:1, 1:10, and 1:100. All

viral samples that were diluted 1:100 were DRP titer around the same concentration as

the cell control tube. This suggests that the samples must contain greater than 107

amount of virus in order to give accurate results without interference with the CMV

probes or background clutter. Wild type AAV2, AAV2 R585A R588A VP1,3 VP2, and

a control were incubated in the harvested Hela cells at 4˚C for 60 minutes allowing for

binding to occur. The samples were washed three times to remove any unbound virus

and the cells were lysed with a lysis buffer. All of the clarified lysates were

Taqman/DRP analyzed for concentrations. Figure 9 demonstrates a linear chart of the

serial dilutions increasing in concentration from left to right. As the amount of virus

increased, the amount of virus recovered for wild type AAV2 also increased. The

heparan mutants continued to stay close to the cell control concentrations. This

concludes that the heparan mutants did not successfully bind to the Hela cells and

appeared similar to background being picked up by Taqman. The cell control should

have came back with an almost zero concentration due to the fact that no virus was

present and thus no CMV promoters were present to be picked up by the CMV probe.

Heparan Cell Binding 1

0

100000000

200000000

300000000

400000000

500000000

600000000

700000000

25000000 250000000 2500000000

Virus added

Vir

us

Bo

un

d

AAV2

HS-VP1,3+VP2Cell lysatewith no virus

Figure 9: Heparan Cell Binding experiment 1. Serial dilutions of the viral preps were from left to right,

1:100, 1:10, and 1:1. As the concentration increased of the viral preps, the amount of cellular binding that did occur also increased. Wild type AAV2 was the only virus that successfully bound to the Hela cells

and the heparan mutants and cell control can back similar to background pickup.

In the second experiment the viral samples were diluted 1:1, 1:10 and 1:50.

Since all viral samples from the first experiment that was diluted 1:100 came back as

just background pickup, a 1:50 dilution was undertaken to see if improvements could be

made. The amount of Hela cells was also increased from previous concentrations of 2.5

x 105 cells to 5.0 x 105 cells so a larger pellet could be seen to control the amount of

cells that was not aspirated off by accident. The viral samples were treated the same as

the first experiment and results are found in Figure 10 below.

Heparan Cell Binding 2

0

10000000

20000000

30000000

40000000

50000000

60000000

70000000

50000000 250000000 2500000000

Virus Added

Vir

us

Bo

un

d

AAV2

HS-VP1,3 +VP2

Cell lysate withno virus

Figure 10: Heparan Cell Binding experiment 2. Serial dilutions were preformed on the viral samples from left to right as 1:50, 1:10, and 1:1. As concluded from the first experiment as the amount of virus

concentration increases, the amount of background pickup decreases and viruses bound increases. AAV2 again was the only significant Hela cell binding when compared to the heparan mutants.

The ability of mutants to bind heparan sulfate on Hela cells was assessed in an

affinity assay in combination with a competitor. This affinity assay will determine the

amount of specific binding that is occurring to the Hela cells as compared to non-

specific binding. Specific binding of wild-type AAV should occur in the absence of a

competitor. AAV binds to its natural receptor HSPG that is expressed on the surface of

Hela cells. AAV should not demonstrate specific binding in the presence of a

competitor in a 100-fold excess that is competing for the same HSPG binding. The

R585A, R588A AAV should demonstrate non-specific binding since its ability to bind

to its natural receptor, HSPG, has been eliminated.

The protocol was set up similar to the cell binding assay, only in conjunction

with the AAV vectors made in Dr. Bartlett’s laboratory with a CMV promoter, another

AAV vector without the CMV promoter was added. The competitor virus called CC119

was a gift from Sferra’s lab, which was a rAAV2 Eflα eGFP at a concentration of 7.6x

1012 DRP/ml. To analyze the concentration of our viral vectors through the

Taqman/DRP titer, a CMV probe is utilized. The CMV probe picked up any virus that

bound to the Hela cells and a concentration would be given. This value is further

calculated out to determine the active concentration of the virus of every particle (DRP)

per microliters.

The viruses were tested to determine if they could bind to the surface of the Hela

cells in the presence of a competitor in 100-fold excess. The samples were allowed to

bind to the Hela cells at 4˚C for 120 minutes. After cell lysis, the lysates was analyzed

by a Taqman/DRP titer to determine the amount of virus in the solution. AAV2 did not

bind as well to the Hela cells similar to the heparan mutants in the cellular binding assay

when the competitor virus was added. Figure 11 demonstrates how AAV2 bound

approximately 32% of the time to the Hela cells in the absence of a competitor. When

competitors to the same heparan binding receptors were added in excess with AAV2,

AAV2 only bound approximately 2% of the time. The heparan mutants really did not

bind either of the times regardless of the competitor.

Figure 11: Hela Cell Binding with competitor. The CC119 competitor was added alongside the viral preps to compete for binding receptors. CC119 is a rAAV2 Eflα eGFP that was added in

100-fold excess of all of the viral preps. Alone, AAV2 bound 32% of the time and AAV2 R585A R588A VP1,3 plus VP2 bound 0.32%. In the presence of a competitor neither virus bound as well.

Hela Cell Binding with Competitor

0

5

10

15

20

25

30

35

40

AAV2 AAV2 + CC119 HS- (VP1,3 +VP2) HS- (VP1,3 +VP2)+ CC119

Perc

en

t B

ou

nd

These preliminary findings suggest that AAV2 does bind to heparan, and that

the rAAV2 mutants were deficient in receptor binding. AAV2 demonstrates specific

binding in the absence of a competitor for the same receptor HSPG. In the presence of a

competitor, AAV2 demonstrates non-specific binding to the Hela cells. The heparan

mutants demonstrated non-specific binding to the Hela cells in both incidences when a

competitor was present and without. This truly demonstrates the elimination of HSPG

binding of the AAV vector. Eliminating the heparan binding portion of the AAV2

vector can lead to successful elimination of unwanted gene transfer and expression and

the retargeting of the rAAV2 vector to tissues that were once thought untargetable.

CHAPTER 5

DISCUSSION AND CONCLUSION

Modifying plasmids with R585A R588A point mutations has successfully

eliminated heparan sulfate proteoglycan binding of the AAV2 vector. Eliminating the

binding to natural receptor of AAV2 and introduction of new peptides improves the

gene transfer and expression of the virus to tissues of interest. AAV2 has the ability to

be the solution to all of the problems of gene therapy. Its ability to invade both dividing

and non-dividing cells, long term gene expression, reproducibility of high titers, and

lack of pathogenicity makes AAV2 a suitable vector for gene therapy. Retargeting

AAV2 to tissues of interests can improve the effectiveness and efficiency of gene

therapy for treating cancers and genetic disorders.

In vitro studies are conclusive that AAV2 with mutations at the R585A R588A

positions does not bind heparin on polystyrene beads, heparan expressing cells, and

specific binding has been eliminated in the presence of a competitor. The R585A

R588A AAV2 mutation does not infect cells, which is thought to be or probably due to

the lack of binding to the heparan expressing Hela cells. What is not apparent is whether

the modified viral vectors can bind to cells expressing their new peptide receptors.

Future studies will be tested on tissue culture cells that have been known to express the

various receptor sites on their surface. VEGF receptor binding ligands were added to the

modified AAV2 vectors deficient in receptor binding. Once tissue tropisms of the AAV

vector are deficient in receptor binding and retargeted to new wanted receptors, in vivo

studies can be investigated.

Future studies will examine the in vivo dissemination of the viral vectors

deficient in natural receptor binding developed in our laboratory. The modified vectors

should transduce cells with matching viral receptors. Gene expression of the modified

heparan mutants will have to be monitored and the levels of the vectors that are

expressed determined. The objectives of the original proposal were to test the modified

heparan mutants mainly in vivo. Goals were set high, but were not unobtainable. As

with any research topic, unforeseen obstacles surfaced. Troubleshooting and pondering

the problems that went wrong helped me to understand the reasoning behind what all I

did. If everything went as planned, I could have completed a lot more objectives

originally proposed, but then it would not be research. Albert Einstein already foresaw

the pitfalls in my research. He once said, “If we knew what it was we were doing, than

it would not be called research, would it?” If research was easy to complete in a short

amount of time, than everyone would be able to fix all of the problems and cancers in

the world. I was given a wonderful opportunity to complete something once believed

unimaginable. With all of my new gained knowledge I acquired from my research, I

feel even more lost in understanding it. As the old saying goes, "The more you know,

the more you realize how much you don't know."

CHAPTER 6

BIBLOGRAPHY

1. Bartlett, J.S., Wilcher, R., Samulski, R. “Infectious Entry Pathway of Adeno-Associated Virus and Adeno-Associated Virus Vectors.” Journal of Virology. 74(6): 2777-2785 (2000).

2. Shi, W., Bartlett, J.S. “RGD Inclusion in VP3 Provides Adeno-Associated Virus Type 2 (AAV2)- Based Vectors with a Heparan Sulfate-Independent Cell Entry Mechanism”. American Society for Gene Therapy: Molecular Therapy. 7-4: 515-525 (2003).

3. Opie, S. R., K. W. Warrington, Jr., et al. “Identification of amino acid residues in the capsid proteins of adeno-associated virus type 2 that contribute to heparan sulfate proteoglycan binding.” Journal of Virology. 77(12): 6995-7006 (2003).

4. Buller, R.M., Rose, J.A.. “Characterization of adenovirus-associated virus – induced polypeptides in KB cells.” Journal of Virology. 25:331-338 (1978).

5. Shi,W., et al. “Insertional mutagenesis of the adeno-associated virus type 2 (AAV2) capsid gene and generation of AAV2 vectors targeted to

alternative cell- surface receptors.” Human Gene Therapy 12:1697-1711(2001).

6. Warrington, K., et al. “ Adeno-associated Virus Type 2 VP2 Capsid Protein Is Nonessential and Can Tolerate Large Peptide Insertions at its N Terminus.” Journal of Virology. 78(12): 6595-6609 (2004).

7. Xie, Q., W. Bu, et al. “The atomic structure of adeno-associated virus (AAV2), a vector for human gene therapy.” Proc Natl Acad Sci USA 99(16): 10405-10 (2002).

8. Kronenberg, S, et al. “Electron cryo-microscopy and image reconstruction of adeno associated virus type 2 empty capsids.” EMBO reports 2(11): 997–1002 (2001).

9. Büchen-Osmond, C. “Dependovirus.” 2002. ICTVdB The Universal Virus Database. 18 March 2005. <http://www.ncbi.nlm.nih.gov/ICTVdb/ICTVdB/index.htm>.

10. Rabinowitz, J., Samulski, R.J. “Adeno-associated virus expression systems for gene transfer”. Current Opinion in Biotechnology.9:470-475 (1998).

11. Feuder, E., Alwis, M., Thrasher, A., Ali, R., Fauser, S. “Optimization of recombinant adeno-associated virus production using an herpes simplex

virus amplicon system.” Journal of Virological Methods. 96, 97-105 (2001).

12. Girod, A., M. Ried, et al. “Genetic capsid modifications allow efficient re-targeting of adeno-associated virus type 2[published erratum appears in Nat Med 1999 Dec; 5(12): 1438].” Nature Medicine 5(9): 1052-6 (1999).

13. Müller, O., et al. “Random peptide libraries displayed on adeno-associated virus to select for targeted gene therapy vectors.” Nature Biotechnology. 21(9): 1040-1046 (2003).

14. Kern, A., K., Schmidt, et al. “Identification of a heparin-binding motif on adeno-associated virus type 2 capsids.” Journal of Virology. 77(20): 11072-11081 (2003).

15. Nicklin, S.A., et al. “Efficient and selective AAV2-mediated gene transfer directed to human vascular endothelial cells.” Mol. Ther. 4(3) : 174-181(2001).

16. Xiao, X., Li, J., and Samulski, R. “Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus.” Journal of Virology. 72: 2224-2232 (1998).

17. Li, J., Samulski, R.j., and Xiao, X. “Role for highly regulated rep gene expression in adeno-associated virus vector production.” Journal of Virology.71: 5236-5243 (1997).

18. Clark, K., Voulgaropoulou, F., and Johnson, P. “A stable cell line carrying adenovirus-inducible rep and cap genes allows for infectivity titration of adeno-associated virus vectors.” Gene Therapy. 3:1124-1132. (1996).

19. Jordan, M., et al. “Transfecting mammalian cells: optimization of critical parameters affecting calcium-phosphate precipitate formation.” Nucleic Acids Res. 24: 596-601(1996).

20. Zolotukhin, S. “Recombinant adeno-associated virus purification using novel methods improves infectious titer and yield.” Gene Therapy 6: 973-985. (1999).

21. Clark, K. R., Liu, X., McGrath, J.P., Johnson, P.R., “Highly purified recombinant adeno-associated virus vectors are biologically active and free of detectable helper and wild-type viruses. Human Gene Therapy 10, 1031-1039.